TABLE 1

Summary of Recent Studies Using Functional Neuroimaging Tools for Examining Pediatric Traumatic Brain Injury

TechniqueSampleResults
Diagnosis and clinical outcome
    Ashwal et al2 (2000)1H-MRS (compared to clinical rating scale results)26 infants (aged 1–18 mo) and 27 children (aged ≥18 mo) with acute nonaccidental traumatic brain injuryMetabolite ratios were abnormal in patients with a poor outcome; lactate was evident in 91% of infants and 80% of children with poor outcomes, and none of the patients with good outcome had lactate; at best, the clinical variables alone predicted the outcome in 77% of infants and 86% of children, and lactate alone predicted the outcome in 96% of the infants and 96% of the children
    Emanuelson et al24 (1997)SPECT (vs CT)20 children (8 girls and 12 boys; mean age: 14.4 y; range: 3.9–18.4) with acquired brain injuriesCT and SPECT revealed similar results in the patients with severe injury, but in the group of mildly injured children the number of affected lobes and scores indicated by SPECT were significantly higher than those indicated by CT; SPECT also differentiated more effectively between the 2 outcome groups
    Goshen et al23 (1996)99Tcm-HMPAO brain SPET (vs MRI and EEG)28 pediatric patients with traumatic brain injury (15 females, 13 males; aged 15 mo to 6 y)Findings indicate that SPET is more sensitive than morphological or electrophysiological tests in detecting functional lesions
    Holshouser et al27 (2000)1H-MRS (comparison of short and long echo time acquisitions)23 acutely brain-injured children (age range: 20 mo to 18 y; mean: 6 ± 4 y, divided into 3 groups [neonates, <1 mo of age; infants, 1–18 mo; children, >18 mo]); causes of brain injury included hypoxic-ischemic encephalopathy, trauma, seizures, cardiac arrest, or near-drowning and cardiac surgeryUsing peak area metabolite ratios and lactate presence, the short- and long-TE methods were equally predictive in children >1 mo of age; in neonates <1 mo of age, the long-TE method produced a higher percentage of correct outcome predictions (91%) than the short-TE method (79%); the long-TE method detected lactate more often in all age groups
    Seghier et al19 (2004)fMRI and DTI3-mo-old infant with unilateral perinatal stroke involving visual pathwaysImages from DTI revealed absence of optic radiation in the damaged hemisphere compared to the intact hemisphere; results from fMRI revealed negative BOLD responses on visual stimulation; supports the combined use of fMRI and DTI to further characterize functional and anatomical consequences of injury to the visual pathways
    Sutton et al26 (1995)1H nuclear magnetic resonance spectroscopy (used to define metabolic abnormalities)15 children (aged 1 mo to 15 y) at the time of head injuryIn comparison with normal brain there was no significant elevation in tissue lactate in the patients with diffuse axonal injury but a significant increase in the regions of contusion and infarction
    Worley et al15 (1995)PET22 previously normal children/adolescents who suffered severe, nonpenetrating traumatic brain injuryPET scores improved significantly between rehabilitation but improvement not associated with improvement in clinical condition; data suggest that routine PET during rehabilitation is no more useful than contemporaneous CT or MRI for prediction of outcome
Neural plasticity and reorganization
    Booth et al49 (2000)fMRI6 pediatric patients with perinatal strokes or periventricular hemorrhages and 7 healthy children (aged 9–12 y)Activation of the intact regions of the damaged hemisphere; left-hemisphere lesion related to organization to homologous areas in the right hemisphere; degree of shift related to size of lesion
    Cioni et al48 (2001)fMRI2 pairs of monozygotic twins; focal brain injury in 1 member of each pairActivation of the undamaged areas adjacent to the lesion site and activation of the ipsilateral sensorimotor cortex in the lesioned subjects; no bilateral activation of the primary sensorimotor cortex in the healthy co-twin controls
    Holloway et al39 (2000)fMRI used to investigate sensorimotor functions of the hand17 children who had undergone hemispherectomy (unilateral brain injury)2 of 8 patients showed activation in the sensorimotor cortex of the remaining hemisphere with passive movement of the hemiplegic hand; location of ipsilateral brain activation similar to that found on movement of the normal contralateral hand, although the latter was greater in spatial extent
    Liégeois et al47 (2004)fMRI10 children (aged 7–18 y) who had intractable epilepsy and were suffering from early left-hemisphere lesions (5 patients sustained lesions within/adjacent to classical language cortex; 5 patients with lesions remote from Broca's or Wernike's area)Right or bilateral language representation was observed in children with left-hemisphere lesions remote from language areas
    Muller et al38 (1998)[15O]water PET21 children with early unilateral brain lesions (2 groups: RL and LL)Number of activated regions overall was greater in the RL group; stronger subcortical and cerebellar language involvement in the RL group; rightward shift of language activations in LL group in perisylvian areas and other temporoparietal regions
    Muller et al34 (1999)PET11 patients with left-hemisphere lesion (aged 8–33 y) and 9 normal adult comparisonsResults suggest enhanced postlesional plasticity in childhood, and also reflects strong individual variability, probably resulting from clinical and demographic factors beside lesion onset
    Niimura et al37 (1999)PET10 children with frontal hypometabolismChildren with lesion onset before 1 y of age showed significantly higher glucose utilization in contralateral posterior quadrangular and superior semilunar lobules of cerebellar cortex than did adults; 5 children showed increased benzodiazepine receptor binding in the dentate nucleus contralateral to lesion; 5 children showed no increase in benzodiazepine receptor binding but showed increased binding in the lateral lobules of the cerebellar cortex contralateral to the lesion
    Stiles et al50 (2003)fMRI2 adolescent subjects (13-y-old male with injury to left hemisphere, 15-y-old male with injury to right hemisphere)Results revealed a contralesional lateralization for global and local pattern information processing
  • RL indicates right lesion; LL, left lesion.